DESCRIPTION (Applicant's Abstract): In clinical settings, it has been demonstrated that African-Americans are more likely to be diagnosed with schizophrenia than Caucasians with a corresponding lower rate of affective disorder diagnoses in the former. In general, when strict diagnostic criteria are applied, these racial differences are not observed, and epidemiologic studies suggest that the rates of major psychoses are not different among racial groups. Misdiagnosis of African-Americans presenting with psychosis may result in increased morbidity and mortality in that patient population. Unfortunately, the factors contributing to this racial discrepancy in diagnosis remain poorly defined. Several investigators have proposed that, compared to Caucasians, African-American patients more commonly experience severe thought disorder, prominent hallucinations and, particularly, the first-rank symptoms of schizophrenia, during the course of both affective and non-affective psychotic disorders. These specific symptoms may cause clinicians to stray from diagnostic criteria leading to an under-recognition of affective syndromes, and excessive clinical diagnoses of schizophrenia in African-American patients. However, few studies have specifically examined this hypothesis prospectively. Moreover, no study has incorporated raters blinded to patient race to guard against rater bias in patient assessments. We propose to examine racial differences in the symptomatic expression of psychosis through the combined use of blinded and unblinded raters to determine whether symptom differences in the expression of first-rank symptoms contribute to racial discrepancies in diagnosis. The novel aspects of this study involve the prospective evaluation of psychosis in African-American and Caucasian patients with psychotic disorders using raters blinded to the patients' race.